Skip to content
Back to skills

Recall

ASecurity

Search VantagePeers for stored memories. Use this skill whenever the user says "recall", "remember", "what do we know about", "search memory", "look up", "find in memory", or asks about past decisions, context, or stored knowledge -- even if they don't say "recall" explicitly.

  • 3 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 29, 2026
ai-agents

Works with

  • mcp

Security analysis

A100/100

Scanned September 29, 2026

npx -y skills add vantageos-agency/vantage-peers --skill recall --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Recall?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Recall
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/vantageos-agency-recall/badge)](https://www.skillsdirectory.com/skills/vantageos-agency-recall)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: recall
description: >
  Search VantagePeers for stored memories. Use this skill whenever the user says
  "recall", "remember", "what do we know about", "search memory", "look up",
  "find in memory", or asks about past decisions, context, or stored knowledge --
  even if they don't say "recall" explicitly.
user-invocable: true
---

Quick semantic search across VantagePeers.

## WORKFLOW

1. Parse the user's query into a search string
2. Determine namespace from context:
   - If the query mentions a project name -> namespace="project/{name}"
   - If the query mentions an agent/orchestrator -> namespace="orchestrator/{agent}"
   - Default -> namespace="global"
3. Call `mcp__vantage-peers__search_memories_by_semantic` query={parsed}, namespace={determined}, limit=5
4. Display results with metadata
5. If results reference related memories, offer to fetch those too

## OUTPUT FORMAT

```
RECALL: "{query}" in {namespace}
Found {n} results:

1. [{type}] {content preview, 2 lines max}
   by {createdBy} on {date} | relations: {count}

2. [{type}] {content preview, 2 lines max}
   by {createdBy} on {date} | relations: {count}
...
```

If no results:
```
RECALL: "{query}" in {namespace}
No results found. Try a different query or namespace.
```

## NAMESPACE DETECTION

Use these heuristics to pick the right namespace:
- User says "about project X" or "in project X" -> namespace="project/x"
- User says "what did pi decide" or "tau's notes" -> namespace="orchestrator/{agent}"
- User says "any feedback about" -> namespace="global" (feedback is always global)
- Ambiguous -> default to "global"

## CONFIGURATION

These can be overridden in CLAUDE.md:
- `default_namespace`: override auto-detection
- `default_limit`: change from 5

## RULES

- Never return raw JSON. Always format for readability.
- Truncate long content to 2 lines with "..." indicator.
- Show relation count so the user can drill deeper.
- All namespace and role values are lowercase.
- If the user asks to "remember" something (store), clarify: "Did you mean to store a memory or search for one?"

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.